Вавиловский журнал генетики и селекции (May 2020)

The use of Specim IQ, a hyperspectral camera, for plant analysis

  • V. V. Alt,
  • T. A. Gurova,
  • O. V. Elkin,
  • D. N. Klimenko,
  • L. V. Maximov,
  • I. A. Pestunov,
  • O. A. Dubrovskaya,
  • M. A. Genaev,
  • T. V. Erst,
  • K. A. Genaev,
  • E. G. Komyshev,
  • V. K. Khlestkin,
  • D. А. Afonnikov

DOI
https://doi.org/10.18699/VJ19.587
Journal volume & issue
Vol. 24, no. 3
pp. 259 – 266

Abstract

Read online

Remote sensing using hyperspectral cameras is an important technology for non-destructive monitoring of plant pigment composition, which is closely related to their physiological state or infection with pathogens. The paper presents the experience of using Specim IQ, a mobile hyperspectral camera, to study common root rot (the pathogen is the fungus Bipolaris sorokiniana Shoem.) affecting the seedlings of four wheat varieties and to analyze the pulp of potato tubers of 82 lines and varieties. Spectral characteristics were obtained for seedlings and the most informative spectral features (indices) for root rot detection were determined based on the data obtained. Seedlings of control variants in the visible part of the spectrum show an increase in reflectance with a small peak in the green area (about 550 nm), then a decrease due to light absorption by plant pigments with an extremum at a wavelength of about 680 nm. Analysis of histograms of vegetation index values demonstrated that the TVI and MCARI indices are the most informative for detecting the pathogen on wheat seedlings according to hyperspectral survey data. For potato samples, regions of the spectrum were found that correspond to local maxima and minima of reflection. It was shown that the spectra of potato varieties have the greatest differences within wavelength ranges of 900-1000 nm and 400-450 nm, which in the former case may be associated with the level of water content, and in the latter, with the formation of melanin in the tubers. It was shown that according to the characteristics of the spectrum, the samples studied are divided into three groups, each characterized by increased or reduced intensity levels for the specified parts of the spectrum. In addition, minima in the reflection spectra corresponding to chlorophyll a were found for a number of varieties. The results demonstrate the capabilities of the Specim IQ camera for conducting hyperspectral analyses of plant objects.

Keywords